Abstract/Summary

TCs have caused death and great economic loss every year across the coastal area of the western North Pacific (WNP). Therefore it is important to improve the understanding of the climatology of TCs over this region and their modulation by natural climate variability and large-scale circulation systems. It is also important to improve our ability to predict possible changes in TC activity over the WNP under climate change conditions. The most appropriate approach to study this is to use numerical models. However, high model resolution is required to resolve the complex physical structure of TCs so that realistic climatologies of TCs can be produced. In this thesis, regional climate model (RCM) simulations, from the Unified Model of the UK Met Office (MetUM) with resolutions of 25 and 12km and two different dynamical cores, are used to project the future change in TC activity over Vietnam, Philippines and the South China Sea (SCS). These simulations are driven by data from the ERA-Interim reanalysis for the period 1990-2005 and the HadGEM2-ES global climate model covering the historical period (1961-2005) and a future period (2069-2099) under two IPCC greenhouse gas emission scenarios (RCP4.5 and RCP8.5) to investigate the impact of anthropogenic warming on TCs in the study region. An objective algorithm is used to identify and track the simulated TCs. First, the ability of the RCM to simulate TCs and their associated large-scale environments, for a current climate period of 1990-2005, with different model resolutions is evaluated. For the period of 1961-2005, the downscaled HadGEM2-ES simulations are also used to evaluate the model's ability to reproduce the modulation of TC activity associated with El Nino/Southern Oscillation (ENSO). Both the 25 km and 12 km models can reasonably simulate the TC activity over the SCS compared with the observed TCs. The associated large-scale environments are found to be simulated correctly compared with the ERA-Interim reanalysis. The observed weakened TC activity during El Nino events is also captured by the downscaled HadGEM2-ES. Compared with the 25 km model, the 12 km model has a better ability to simulate the large-scale environments and generally improves the simulation of the spatial distribution and structure of TCs. Improved simulated TC-ENSO response (in terms of TC frequency, track density, intensity, structure and associated large-scale environments) is also found in the 12 km model. However, the 12 km model does not produce stronger 10-m maximum wind speeds of TCs compared with the 25km model. For the projections towards the end of the 21st century, the downscaled HadGEM2-ES at both 25 km and 12 km resolution present an insignificant decrease in TC frequency with rates of -0.001 per year and -0.007 per year for RCP4.5 and RCP8.5 respectively. The Had2-25km projection shows a more than 3% increase of the intense (10 m wind speed >35 mS-1) TCs over the South China Sea, while weak TCs (wind speed >25 ms-1) decrease by 10% under RCP8.5. Also, both the RCMs simulate a seasonal shift of TC activity in a warming climate, with an increase in TCs during winter related to the more favourable large-scale conditions and a decrease in TCs is projected in summer.